Blood Lipid Levels and Intraocular Pressure in a Japanese Population: the JPHC-NEXT Eye Study

Abstract

Aim: Although growing epidemiological evidence suggests that metabolic risk factors may influence the intraocular pressure (IOP), the specific effects of individual lipid components on IOP regulation remain unclear. In this study, we examined the associations between the serum lipid profiles, individually and in combination, and IOP among 6,786 ophthalmologically healthy adults aged ≥ 40 years residing in Chikusei City, Japan.

Methods: Participants with a history of glaucoma, the use of IOP-lowering medications, or previous ocular surgery were excluded. Comprehensive systemic and ophthalmological evaluations, including blood tests and measurements of IOP and central corneal thickness, were performed. An analysis of covariance was used to assess the associations of low-density lipoprotein cholesterol (LDL-C), high-density lipoprotein cholesterol (HDL-C), and triglyceride (TG) levels with IOP.

Results: After adjusting for several potential confounders, higher LDL-C, HDL-C, and TG levels were significantly associated with increased IOP. Specifically, every 10 mg/dL increase in LDL-C, HDL-C, and TG corresponded to IOP increases of 0.04±0.011 mmHg, 0.13±0.026 mmHg, and 0.02±0.006 mmHg, respectively. Notably, the positive association between LDL-C and IOP was more pronounced in individuals aged ≥ 65 years than in those aged ＜65 years (P for interaction = 0.009).

Conclusion: These findings suggest that lipid homeostasis may contribute to IOP regulation, particularly in older populations.

See editorial vol. 33: 1259-1260

Introduction

Glaucoma, characterized by progressive optic neuropathy and distinctive visual field loss, has become the leading cause of irreversible blindness worldwide ^{1 , 2)} . Elevated intraocular pressure (IOP) is a primary factor contributing to glaucomatous optic nerve damage and subsequent visual field deficits ³⁾ leading to substantial limitations in daily functioning and a loss of autonomy ⁴⁾ . The only available treatment is IOP reduction. Methods for lowering IOP include laser treatment, topical medication, and surgery. The mechanism underlying IOP regulation remains unclear, highlighting the need to investigate the relationship between systemic conditions and IOP to better understand these mechanisms.

Epidemiological studies have suggested that IOP is influenced by several factors, including non-modifiable risk factors such as age ⁵⁾ , race ⁵⁾ , refraction ⁶⁾ , and central corneal thickness (CCT) ⁷⁾ , as well as modifiable risk factors such as blood pressure ⁸⁾ , diabetic status ⁹⁾ , physical activity ¹⁰⁾ , and obesity ¹⁰⁾ . Lifestyle-related diseases, such as hypertension, diabetes, and dyslipidemia, contribute to arteriosclerosis progression, which may play a role in increasing IOP. Arteriosclerosis leads to reduced vascular elasticity and an impaired blood flow, which subsequently affects the intraocular circulation and ultimately results in elevated IOP ¹¹⁾ . Among the metabolic risk factors, a large-scale study of the Japanese population has reported a positive association between hypertension and diabetes with IOP ⁸⁾ .

Lipids are classified into low-density lipoprotein cholesterol (LDL-C), high-density lipoprotein cholesterol (HDL-C), and triglycerides (TG), with each type of lipid having a different impact on IOP ¹²⁾ . Several studies ^{13 , 14)} showed positive associations between serum lipid levels (total cholesterol, HDL-C, and TG) and IOP. A meta-analysis revealed an inverse association between HDL-C and glaucoma ¹⁵⁾ . More recently, Madjedi et al. analyzed large cohorts from the UK Biobank and EPIC-Norfolk studies and showed that higher levels of LDL-C, HDL-C, and total cholesterol were positively associated with IOP, whereas TG showed an inverse association with IOP in the UK Biobank ¹⁶⁾ . Taken together, these prior studies indicate that serum lipids may contribute to IOP regulation; however, most large cohort data have been derived from European populations, and there have been a limited number of epidemiological studies examining lipid levels among East Asian populations ¹³⁾ . Lipid metabolism, as well as glaucoma subtypes, is known to differ across ethnic groups. For example, large multiethnic cohorts have shown distinct distributions of LDL-C, HDL-C, and TG in European, East Asian, and South Asian populations ¹⁷⁾ . In addition, East Asian populations have a higher prevalence of normal-tension glaucoma, whereas high-tension open-angle glaucoma is more common in European populations. Therefore, extending previous findings from predominantly European cohorts to an East Asian population is important for understanding whether lipid-IOP association is consistent across ethnicities. Furthermore, few studies have investigated differential associations by lipid type, while accounting for other metabolic factors, including hypertension and diabetes. Therefore, we conducted a population-based study within the JPHC-NEXT Eye Study, to investigate the associations between multiple serum lipid traits and IOP in a large Japanese cohort.

Aim

We aimed to examine the associations between the lipid levels and IOP after comprehensively adjusting for potential confounding factors in a large number of relatively healthy Japanese populations.

Methods

Study Population

The Japan Public Health Center-based Prospective Study for the Next Generation (JPHC-NEXT) is an ongoing population-based cohort study designed to investigate the lifestyle and environmental factors of chronic diseases in several regions of Japan ¹⁸⁾ . Participants in the JPHC-NEXT Eye Study were primarily recruited from the Chikusei area of the JPHC-NEXT Study and comprised health checkup examinees aged ≥ 40 years who provided informed consent to participate in the Eye Study. Participants were interviewed by trained technicians or completed a self-administered questionnaire that inquired about their medical history, including dyslipidemia, diabetes, and hypertension, as well as histories of ocular diseases and related surgeries, smoking and alcohol consumption habits, and medication use. Participants also underwent systemic and ophthalmological examinations.

Among the 9,939 participants who consented to participate in the survey and were aged ＞40 years, we excluded those diagnosed with glaucoma or with a history of IOP-lowering treatment (n = 491). We also excluded those with a history of ocular surgery (n = 1,432), incomplete data on anthropometric or laboratory measurements (n = 1,069), and individuals with missing or implausible IOP measurements (IOP ＜5 or IOP ＞40 mmHg) (n = 618). Finally, 6,786 participants (2,771 men and 4,015 women) were included in the analysis.

Written informed consent was obtained from all the participants. The study adhered to the Ethical Guidelines for Medical and Health Research Involving Human Subjects in Japan and was approved by the Medical Ethics Committees of Keio University (Tokyo), University of Tsukuba (Ibaraki), Osaka University (Osaka), and National Cancer Center (Tokyo).

Systemic Examination

Anthropometric parameters, such as waist circumference, weight, and height, were measured by trained technicians using standard protocols. The body mass index (BMI) was calculated by dividing the weight (kg) by the square of the height (m). Blood pressure measurements were taken twice on the right upper arm while the participant was seated, and the average of these two readings was used for analysis. Blood samples were collected to measure serum lipid levels, including LDL-C (mg/dL), HDL-C (mg/dL), TG (mg/dL), and glycated hemoglobin (HbA1c; %). The participants were examined in a fasting state, although this was not mandatory. Hypertension was defined as the use of antihypertensive medication or systolic blood pressure ≥ 140 mmHg and/or diastolic blood pressure ≥ 90 mmHg. Diabetes was defined as the use of antidiabetic medication or HbA1c level ≥ 6.5%.

Ascertainment of Exposure

Dyslipidemia was defined as the use of any antihyperlipidemic medication or LDL-C ≥ 140 mg/dL, and/or HDL-C ＜40 mg/dL, and/or TG ≥ 150 mg/dL (fasting state) or ≥ 175 mg/dL (non-fasting state) levels, in accordance with the Japanese guidelines for the prevention of atherosclerotic cardiovascular diseases ¹⁹⁾ . In the main analysis, the LDL-C, HDL-C, and TG levels were categorized into quintiles for the analysis.

Ocular Examination and IOP Measurement

Ocular examinations, including IOP and CCT measurements, were performed. The IOP of both eyes was measured using a non-contact tonometer (Tonoref™ II; Nidek Co., Ltd., Aichi, Japan), and the mean value of three measurements from both eyes was used for this analysis. The CCT of both eyes was measured using a specular type pachymeter (Specular Microscope XIII; Konan, Nishinomiya, Japan). Ocular hypertension was defined as an average IOP ＞21 mmHg in both eyes, excluding cases of optic disc abnormality.

Statistical Analysis

The association between serum lipids and IOP was assessed using two models of an analysis of covariance. In our multivariate model, the first model was adjusted for age (continuous) and sex (male vs. female). The second model included additional adjustments for smoking status (never, former, or current), alcohol intake (never, occasional, or current), hypertension (yes or no), diabetes (yes or no), BMI (＜25 vs. ≥ 25 kg/m²), CCT (continuous), antihyperlipidemic medication (yes or no), and serum lipids (LDL-C, HDL-C, and TG, excluding the exposure lipids). Changes in IOP across lipid quintiles were calculated for both age- and sex-adjusted models and multivariable-adjusted models. Additionally, the IOP per 10 mg/dL increase in lipids was estimated.

Moreover, we examined the combined effect of each lipid profile on the IOP. Each lipid profile was categorized using the median value of the respective lipid parameter (124.0 mg/dL for LDL-C, 62.0 mg/dL for HDL-C, and 96.0 mg/dL for TG, regardless of fasting status), classifying participants into eight groups. For instance, those with lower LDL-C, HDL-C, and TG levels were categorized as Group 0, while those with higher levels were categorized as Group 7. We then assessed whether the IOP differed according to the group classification.

In an exploratory analysis, we further examined whether the observed positive relationship between lipid profiles and IOP differed by age (＜65 vs. ≥ 65 years), sex (men vs. women), and the presence of hypertension or diabetes (yes vs. no). Specifically, we used ＜65 years, men, and individuals without hypertension or diabetes as reference groups to compare the strength of the associations in the other subgroups. The statistical significance of the multiplicative interaction was evaluated by applying the Wald test to the product term of each lipid profile (10 mg/dL) and the stratification variables. All P-values were two-sided, with a significance threshold set at P＜0.05. Analyses were performed using the SAS software program (version 9.4; SAS Institute Inc., Cary, NC, USA).

Results

Baseline Characteristics of Participants According to the Lipidemic Status

Of the 6,786 participants, 2,834 (40.9%, comprising 1,159 men and 1,675 women) were identified as having dyslipidemia. Table 1 presents the baseline characteristics of the participants stratified by dyslipidemia. Compared to participants without dyslipidemia, those with dyslipidemia were older and more likely to have a higher BMI. In addition, a history of diabetes and hypertension was more frequently observed in individuals with dyslipidemia. Furthermore, the mean IOP±standard deviation was higher in participants with dyslipidemia (14.0±2.8 mmHg) than in those without (13.8±2.8 mmHg). The distribution of IOP according to the lipidemic status is shown in Fig.1 . The mean values and standard errors for LDL-C, HDL-C, and TG were 126.01±0.40 mg/dL, 63.4±0.20 mg/dL, and 113.36±0.89 mg/dL, respectively.

Table 1. Baseline Demographic and Clinical Features According to Dyslipidemia^＊ .

Characteristics	Non-dyslipidemia (N = 3952)	Dyslipidemia‡ (N = 2834)	P-value
Age in years (SD)	62.7 (10.5)	65.1 (8.7)	＜0.0001
Men, n (%)	40.7	40.9	0.90
Body mass index, kg/m2 (SD)	22.6 (3.2)	24.1 (3.2)	＜0.0001
LDL cholesterol, mg/dL (SD)	123.6 (29.6)	129.5 (33.6)	＜0.0001
HDL cholesterol, mg/dL (SD)	66.7 (15.3)	58.3 (15.2)	＜0.0001
Triglyceride, mg/dL (SD)	85.8 (30.2)	152.6 (88.3)	＜0.0001
Total cholesterol, mg/dL (SD)	206.7 (33.1)	214.0 (37.7)	＜0.0001
HbA1c, % (SD)	5.7 (0.6)	5.9 (0.7)	＜0.0001
Waist circumference, cm (SD)	73.7 (24.9)	76.9 (27.1)	＜0.0001
Systolic blood pressure, mmHg (SD)	125.9 (17.7)	130.7 (17.0)	＜0.0001
Diastolic blood pressure, mmHg (SD)	74.6 (11.5)	76.6 (11.0)	＜0.0001
Intraocular pressure, mmHg (SD)	13.8 (2.8)	14.0 (2.8)	0.04
Central corneal thickness, μm (SD)	551.4 (58.6)	549.1 (58.1)	0.14
Hypertension, n (%)¶	42	52.8	＜0.0001
Diabetes, n (%)§	8.0	15.2	＜0.0001
Medication for hypertension, n (%)	23.5	41.4	＜0.0001
Medication for diabetes, n (%)	5.3	9.6	＜0.0001
Alcohol intake (%)			＜0.0001
23 g/day	54.8	62.2
23 to ＜46 g/day	17.3	15.4
≥ 46 g/day	27.9	22.4
Smoking (%)			0.34
Never	62	61.0
Former	25.6	25.8
Current	12.5	13.2
Fasting status (%)			＜0.0001
Non-fasting	24.5	44.3
Fasting	75.5	55.7

^＊Values are presented as the means (SDs) for continuous variables and percentages for categorical variables.

^＊＊Dyslipidemia was defined as LDL cholesterol ≥ 140 mg/dL or HDL cholesterol ＜40 mg/dL or Triglyceride ≥ 150 mg/dL (fasting) or ≥ 175 mg/ dL (non-fasting) or self-reported dyslipidemia medication use.

^＊＊＊Hypertension was defined as either the use of antihypertensive medication or having systolic blood pressure ≥ 140 mmHg and/or diastolic blood pressure ≥ 90 mmHg.

^§Diabetes was defined as either the use of antidiabetic medication or an HbA1c ≥ 6.5%.

Abbreviations: HDL, high-density lipoprotein; LDL, low-density lipoprotein; HbA1c, glycated hemoglobin A1c; SD, standard deviation.

Fig.1.

Distribution of the Intraocular Pressure According to the Dyslipidemia Status

Associations of LDL-C, HDL-C, and TG with IOP

Table 2 presents the associations between each lipid profile categorized into quintiles. We used quintiles to create groups of comparable sizes and thereby reduce sparse-data bias ^{20 , 21)} . We did not apply fixed numeric cut points because no universally accepted categorization exists for these biomarkers in relation to IOP, and guideline thresholds vary across populations and contexts. After adjusting for several confounding factors, a 10 mg/dL increase in LDL-C, HDL-C, and TG corresponded to an increase in IOP of 0.041 mmHg, 0.13 mmHg, and 0.022 mmHg, respectively ( Table 2 ) . When each serum lipid level was classified into quintiles, a statistically significant trend toward increased IOP was observed in participants with higher categories of each serum lipid level (P_trend ＜0.01) ( Table 2 ) .

Table 2. Adjusted Mean Intraocular Pressure According to Serum Lipid Levels.

	n	Mean Intraocular Pressure, mmHg (SE)
		Age and sex adjusted		Multivariable adjusted†
		Means	Ptrend	Means	Ptrend
LDL-C, mg/dL			0.0017		＜0.0001
29-99	1383	13.68 (0.08)		13.65 (0.078)
100-116	1302	13.73 (0.078)		13.76 (0.075)
117-132	1407	14.08 (0.078)		14.04 (0.074)
133-150	1311	13.88 (0.08)		13.89 (0.077)
151-332	1383	14.08 (0.079)		14.11 (0.077)
10 mg/dL		+0.039 (0.011)		+0.041 (0.011)
HDL-C, mg/dL			0.57		＜0.0001
17-49	1368	13.81 (0.082)		13.61 (0.086)
50-57	1308	13.9 (0.08)		13.81 (0.078)
58-65	1381	14.03 (0.079)		14 (0.075)
66-75	1348	13.81 (0.079)		13.9 (0.077)
76-164	1381	13.91 (0.08)		14.15 (0.081)
10 mg/dL		+0.003 (0.002)		+0.13 (0.026)
TG, mg/dL			＜0.0001		0.0015
23-64	1325	13.57 (0.078)		13.65 (0.083)
65-84	1411	13.89 (0.080)		13.88 (0.077)
85-109	1355	13.90 (0.079)		13.90 (0.075)
110-148	1337	13.96 (0.079)		13.93 (0.077)
149-1206	1358	14.14 (0.080)		14.10 (0.082)
10 mg/dL		+0.026 (0.005)		+0.022 (0.006)

^＊Adjusted for age, sex, BMI, diabetic, hypertensive, smoking, alcohol, fasting status, CCT, antihyperlipidemic medication, LDL-C, HDL-C, and TG.

Hypertension was defined as either the use of antihypertensive medication or having systolic blood pressure ≥ 140 mmHg and/or diastolic blood pressure ≥ 90 mmHg. Diabetes was defined as either the use of antidiabetic medication or an HbA1c ≥ 6.5%.

Abbreviations: BMI, body mass index; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; SE, standard error; TG, triglyceride; CCT, central corneal thickness.

IOP Differences among Groups Based on the Median Levels of LDL-C, HDL-C, and TG

The participants were divided into eight groups based on whether their LDL-C, HDL-C, and TG levels were above or below the respective medians ( Table 3 ) . The analysis revealed that the group with all lipid levels above the median had the highest mean IOP (14.2±0.11), while the group with all lipid levels below the median exhibited the lowest mean IOP (13.4±0.12).

Table 3. IOP among Groups Based on Combinations of Each Lipid Level.

Group	LDL-C	HDL-C	TG	N	Mean IOP, mmHg (SE)
0	Low	Low	Low	547	13.4 (0.12)
1	Low	High	Low	1153	13.9 (0.08)
2	Low	Low	High	899	13.7 (0.09)
3	Low	High	High	419	13.8 (0.13)
4	High	Low	Low	416	13.7 (0.14)
5	High	High	Low	900	14.0 (0.09)
6	High	Low	High	1192	14.1 (0.08)
7	High	High	High	609	14.2 (0.11)

We dichotomized the population for “High” vs. “Low” by using the median value of each serum lipid: 124.0 mg/dL for LDL-C, 62.0 mg/dL for HDL-C, and 96.0 mg/dL for TG, respectively.

Abbreviations: IOP, intraocular pressure; HDL, high-density lipoprotein; LDL, low-density lipoprotein; SE, standard error; TG, triglyceride

Stratified Analysis of the Lipid Levels with IOP according to Age, Sex, the Presence of Hypertension or Diabetes, and the Use of Antihyperlipidemic Medications

We conducted stratified analyses based on age, sex, presence of hypertension or diabetes, and use of hyperlipidemic medications ( Supplementary Table 1 ) . The positive association between LDL-C and IOP was stronger in older participants than in younger individuals (P_interaction = 0.009). A non-significant positive interaction was also observed between TG and IOP. In contrast, the positive association between lipid profiles and IOP did not differ by sex, hypertension, or diabetes (P _interaction ＞0.05).

Supplementary Table 1. Interaction Effects of Lipid Levels, Hypertension, Diabetes, Age, Sex, and Antihyperlipidemic Medication Use on Intraocular Pressure (IOP).

	Change in IOP per 10 mg/dL increase in each lipid＊, †
	LDL-C	HDL-C	Triglyceride
Normotensive (n = 2815)	Reference	Reference	Reference
Hypertensive‡ (n = 3971)	0.035 (-0.007, 0.077)	0.065 (-0.021, 0.150)	0.002 (-0.018, 0.022)
	Pinteraction = 0.11	Pinteraction = 0.56	Pinteraction = 0.85
Non-diabetes (n = 6058)	Reference	Reference	Reference
Diabetes§ (n = 728)	0.014 (-0.052, 0.081)	0.016 (-0.012, 0.044)	0.016 (-0.012, 0.044)
	Pinteraction = 0.85	Pinteraction= 0.85	Pinteraction = 0.56
Age ＜65 (n = 3152)	Reference	Reference	Reference
Age ≥ 65 (n = 3634)	0.057 (0.014, 0.099)	0.018 (-0.067, 0.103)	-0.019 (-0.038, 0.0007)
	Pinteraction = 0.009	Pinteraction = 0.68	Pinteraction = 0.059
Men (n = 2771)	Reference	Reference	Reference
Women (n = 4015)	-0.019 (-0.062, 0.025)	-0.043 (-0.134, 0.048)	-0.0057 (-0.025, 0.014)
	Pinteraction = 0.40	Pinteraction = 0.35	Pinteraction = 0.57
No antihyperlipidemic medication (n = 5314)	Reference	Reference	Reference
Antihyperlipidemic medication (n = 1472)	0.040 (-0.020, 0.101)	0.101 (-0.014, 0.216)	0.022 (-0.005, 0.049)
	Pinteraction = 0.91	Pinteraction = 0.54	Pinteraction = 0.93

^＊Changes in IOP per 10 mg/dL increase in each lipid relative to the reference are shown in the lower row (95% confidence interval).

^†Adjusted for age, sex, BMI, diabetic, hypertensive, smoking, alcohol, fasting status, CCT, and serum lipids (LDL-C, HDL-C, and TG, excluding the exposure lipids)

^‡Hypertension was defined as either the use of antihypertensive medication or having systolic blood pressure ≥140 mmHg and/or diastolic blood pressure ≥ 90 mmHg.

^§Diabetes was defined as either the use of antidiabetic medication or an HbA1c ≥ 6.5%.

Abbreviations: BMI, body mass index; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; CCT, central corneal thickness; TG, triglyceride; HbA1c, glycated hemoglobin

We also examined the influence of antihyperlipidemic medications on the associations between serum lipid levels and IOP. In stratified analyses according to antihyperlipidemic medication use, higher LDL-C, HDL-C, and TG levels were each positively associated with IOP among the participants not taking antihyperlipidemic medications (n = 5,314). In this group, a 10 mg/dL increase in LDL-C, HDL-C, and TG was associated with increases in IOP of 0.044 mmHg (95%CI: 0.019−0.068), 0.13 mmHg (95%CI: 0.083−0.187), and 0.02 mmHg (95%CI: 0.010−0.032), respectively. Among the participants taking antihyperlipidemic medications (n = 1,472), the corresponding estimates were 0.040 mmHg (95%CI: -0.020−0.101) for LDL-C, 0.101 mmHg (95%CI: -0.014−0.216) for HDL-C, and 0.022 mmHg (95%CI: -0.005−0.049) for TG. Although these associations were not statistically significant in the medication group, their directions were consistent with those observed in the non-medication group, suggesting that the lack of statistical significance may be attributable to the smaller sample size. The interaction terms between each lipid trait and antihyperlipidemic medication use were not statistically significant (all P for interaction ＞0.05).

Discussion

In this large cross-sectional study, we found positive associations between the LDL-C, HDL-C, and TG levels and IOP. In our sample, a 10 mg/dL increase in LDL-C, HDL-C, and TG corresponded to increases in IOP of 0.04 mmHg, 0.13 mmHg, and 0.02 mmHg, respectively. Furthermore, our combined analyses of each lipid profile showed the highest IOP in the group in which all lipid profiles were above the median value. These results suggest the importance of managing different lipid profiles to control IOP, particularly in older populations.

Although the IOP increase associated with higher lipid levels is small and may have limited immediate clinical significance at the individual level, a population-based prevention strategy aimed at lowering the mean level of IOP and shifting the whole distribution of IOP in a favorable direction could reduce the number of people with high IOP and potentially decrease the number of people with POAG in the population ²²⁾ . Our results were generally comparable to those reported in previous meta-analyses, which indicated that a 10 mg/dL increase in LDL-C, HDL-C, and TG corresponded to increases in IOP by 0.050 mmHg, 0.022 mmHg, and 0.016 mmHg, respectively ²³⁾ . In addition, we compared the effect sizes of the serum lipid levels with those of other systemic factors associated with IOP in the same cohort. In our previous analyses of the JPHC-NEXT Eye Study ⁸⁾ each 10 mmHg increase in systolic blood pressure was associated with a 0.32 mmHg increase in the IOP. In the present cohort, the SDs of systolic blood pressure, LDL-C, HDL-C, and TG were 17.7 mmHg, 29.6 mg/dL, 15.3 mg/dL, and 30.2 mg/dL, respectively. Thus, a 0.56-SD increase in LDL-C, HDL-C, and TG levels would be expected to increase IOP by 0.07 mmHg, 0.11 mmHg, and 0.03 mmHg, respectively, whereas a 0.56-SD increase in systolic blood pressure was associated with a 0.32 mmHg increase in IOP. Similarly, in the JPHC-NEXT Eye Study, the mean IOP was 0.5 mmHg higher in participants with diabetes than in those without diabetes ⁹⁾ , whereas in the present study, the difference in mean IOP between participants with and without dyslipidemia was approximately 0.2 mmHg. These comparisons indicate that the influence of serum lipid levels on IOP is smaller than that of blood pressure and diabetes in this population.

Madjedi et al. reported an inverse association between the TG levels and IOP in the UK Biobank cohort. However, this finding was not observed in the EPIC-Norfolk cohort. In our study, the TG levels were positively associated with IOP. However, the effect size was small, and previous meta-analyses have also indicated only a minimal change in IOP per 10 mg/dL difference in TG levels. Taken together, these observations suggest that the association between TG and IOP is weak and may vary across studies, possibly reflecting differences in study populations and designs. In addition, TG levels are influenced by diet, alcohol intake, and other lifestyle factors and correlate with other metabolic factors, such as LDL-C and BMI, which may contribute to the instability of the results. Moreover, TG levels vary with fasting status and the time since the last meal. Therefore, including a mixture of fasting and non-fasting measurements may increase the variability in the observed TG-IOP associations.

The pathogenesis underlying the association between lipids and IOP remains unclear. However, the LDL-C and TG levels are likely to affect IOP through mechanisms related to outflow resistance. Previous studies have reported that elevated TG levels are associated with increased IOP and a higher risk of primary open-angle glaucoma (POAG) ^{14 , 24)} . Several hypotheses suggest that hypertriglyceridemia may lead to increased blood viscosity, which in turn enhances episcleral venous pressure, resulting in increased resistance to an aqueous humor outflow and subsequently increasing IOP ¹²⁾ . Furthermore, increased blood viscosity may reduce optic nerve blood flow ²⁵⁾ which is considered another contributing factor. A previous study showed that patients with POAG have lower optic nerve blood flow, which may be related to the higher blood viscosity in these individuals ²⁵⁾ .

The mechanism by which LDL-C increases IOP appears to be similar to that of TG, in which elevated LDL-C levels increase blood viscosity ²⁶⁾ . In addition, LDL-C is susceptible to oxidation, and oxidized LDL-C can damage endothelial cells in blood vessels ²⁷⁾ , leading to the formation of atherosclerotic plaques, vascular dysfunction ²⁸⁾ , and an increase in IOP. An interaction analysis revealed that an increase in LDL-C was associated with a significant rise in IOP among individuals aged ≥ 65 years compared to those aged ＜65 years.

This may be explained by the fact that aging is accompanied by increased oxidative stress on the vasculature ²⁸⁾ , making older adults more susceptible to the effects of dyslipidemia, which in turn may lead to elevated IOP. Additionally, older adults may have been exposed to dyslipidemia for a longer duration since their youth, leading to arteriosclerotic vascular changes that could contribute to a higher IOP in later life.

We observed a positive association between total HDL-C levels and IOP, which appears counterintuitive, as HDL-C is generally regarded as a protective lipid against atherosclerotic changes ²⁹⁾ . However, the mechanisms by which lipid traits influence IOP are likely to be more complex than atherosclerosis alone. IOP is determined not only by ocular perfusion and vascular stiffness but also by aqueous humor dynamics and episcleral venous pressure. In addition, the HDL-C levels are influenced by lifestyle and metabolic factors, including alcohol intake ³⁰⁾ , physical activity ³¹⁾ , and obesity ³²⁾ which may explain why observational studies sometimes report positive rather than inverse associations between HDL-C and IOP. Typically, HDL-C protects tissues by reducing inflammation and oxidative damage. In addition, HDL-C comprises various subclasses, each with distinct neuroprotective effects. Among these subclasses, HDL3-C has been reported to significantly reduce the risk of POAG ¹²⁾ . A two-sample Mendelian randomization study using genome-wide association data revealed that higher levels of HDL3-C cholesterol were associated with decreased odds of POAG (odds ratio, 0.91; 95% confidence interval, 0.84–0.99, P = 0.021) ¹²⁾ . Notably, routine lipid measurements, including LDL-C and total HDL-C, were not associated with POAG ¹²⁾ . The observation that individuals with POAG have lower levels of HDL3-C implies less efficient cholesterol transport through the reverse cholesterol transport pathway ³³⁾ . A decrease in HDL3-C level could hinder the removal of cholesterol from tissues to the bloodstream, resulting in increased cholesterol accumulation in tissues, including the retina ³³⁾ . However, cholesterol is an essential component of neuronal physiology, and maintaining its homeostasis in the retina is crucial to avoid toxic accumulation. Consequently, future studies are necessary to investigate the association between more detailed qualitative HDL-C subclasses and IOP.

In our stratified analyses, the positive association between dyslipidemia and IOP remained consistent, irrespective of the presence of hypertension or diabetes. This consistency suggests that the association between dyslipidemia and IOP may operate independently of these common comorbid conditions, highlighting a potentially distinct pathophysiological mechanism ^{13 , 26)} . In patients with hypertension, increased blood pressure may lead to increased ultrafiltration of aqueous humor due to elevated ciliary artery pressure in the ciliary body and elevated episcleral venous pressure, which affects the aqueous humor outflow, ultimately resulting in increased IOP ³⁴⁾ . In patients with diabetes, the accumulation of advanced glycation end products in trabecular meshwork cells induces oxidative stress and apoptosis ^{35 , 36)} which may impair the trabecular meshwork function and contribute to an elevated IOP ^{28 , 37)} . Given that IOP tends to increase with the number of metabolic syndrome components exhibited by an individual ³⁸⁾ , it can be inferred that each component may elevate IOP through different mechanisms. Future studies are warranted to investigate which populations might benefit from managing the lipid metabolism to control IOP.

Dyslipidemia can contribute to changes in IOP and the development of glaucoma through multiple pathways. Whether therapeutic lipid interventions can modify IOP or glaucoma risk is an important aspect. The potential influence of antihyperlipidemic medications, particularly statins, warrants consideration. Statins are lipid-lowering agents that inhibit HMG-CoA reductase, thereby suppressing hepatic cholesterol synthesis and enhancing the uptake of circulating LDL-C via LDL receptors. A recent systematic review and meta-analysis by Yuan et al. reported that statin use was associated with a modest reduction in the risk of POAG onset, whereas it was not clearly associated with an IOP-lowering effect ³⁹⁾ . Statins are thought to reduce the risk of glaucoma by decreasing aqueous outflow resistance and increasing ocular perfusion through enhanced nitric oxide production in vascular endothelial cells and remodeling of the extracellular matrix, as well as by exerting neuroprotective effects through anti-inflammatory actions ⁴⁰⁾ . The findings from the meta-analyses suggest that statins reduce the risk of glaucoma through pleiotropic effects, such as neuroprotection, rather than by lowering IOP ^{40 , 41)} .

The strengths of this study include its large sample size, the use of questionnaires to gather information regarding lifestyle and medical history, and the adjustment for major confounders of the examined associations. Moreover, experienced ophthalmologists conducted ocular evaluations based on a thorough examination. However, this study has several limitations. First, owing to the cross-sectional nature of our study, we could not draw any conclusions regarding causality. Nonetheless, few population-based studies have been conducted on East Asians, making our research important despite these limitations. Furthermore, IOP was measured using a non-contact tonometer instead of the Goldmann applanation tonometer (considered the most precise and reliable tool for measuring IOP). However, in this large population-based setting, a non-contact tonometer was considered a pragmatic choice.

In conclusion, the LDL-C, HDL-C, and TG levels were positively associated with IOP in an ophthalmologically healthy Japanese population. Although this association was particularly evident among older adults, who may be more vulnerable to lipid-related vascular changes due to increased oxidative stress and prolonged exposure to dyslipidemia over their lifetime, this likely reflects the cumulative burden of dyslipidemia over the lifespan rather than changes occurring only in older age. Importantly, the absolute IOP changes per unit increase in lipid levels were small, suggesting limited immediate clinical significance for individual patients. Nevertheless, clarifying how lipids influence IOP may be valuable for elucidating the mechanisms underlying IOP regulation and glaucoma pathogenesis. Therefore, maintaining a healthy lipid profile from a young age may be crucial for preventing IOP elevation and potentially reducing the risk of glaucoma later in life.

Conflicts of Interest

All authors certify that they have no financial relationships or other conflicts of interest to disclose related to the subject matter of this manuscript.

Declaration of Generative AI in Scientific Writing

The authors declare that AI was not used to generate the content in this article.

Acknowledgment

We thank all participants and the staff of the JPHC-NEXT protocol for their valuable contributions.

Financial Support

This work was supported by the National Cancer Center Research and Development Fund (since 2011) and a Grant-in-Aid for Scientific Research from the Japan Society for the Promotion of Science, Japan [JP23K09732 and JP25K20188]. The funders had no role in the design or conduct of the study; collection, management, analysis, or interpretation of the data; preparation, review, or approval of the manuscript; or the decision to submit the manuscript for publication.